JP7250155B2 - Fish counting system, fish counting method and program - Google Patents

Description

The present disclosure relates to fish counting systems, fish counting methods and programs.

In aquaculture of fish such as tuna, it is preferable to accurately grasp the number of fish, reduce feed costs, and plan production. Conventionally, the fish are counted by moving the fish from the first cage to the second cage and photographing the moving fish. Since the number of fish was counted visually while playing the recorded video in slow motion, the number of man-hours increased and accuracy was lacking.

Patent Literature 1 discloses a system that counts passing fish using a plurality of cameras in order to count fish passing through a fishway such as a river.

JP 2005-202674 A

However, the accuracy of counting by a camera may vary depending on changes in the imaging environment such as the direction of the sun and turbidity of water. Also, when a large number of fish move together, an erroneous counting may occur due to duplication of fish.

Therefore, it is preferable to be able to manually correct the counting results, as erroneous counting may occur. However, it is cumbersome to visually check all moving images taken for a long period of time to check whether or not there is an erroneous counting, which increases the number of man-hours.

The present disclosure provides a fish counting system, a fish counting method, and a program that can easily correct counting errors.

The fish counting system of the present disclosure includes an image acquisition unit that acquires a plurality of images of a fluid containing fish captured over time, a counting unit that counts the number of fish based on the plurality of images, and A fish count change display providing unit that provides a fish count change display in which displays corresponding to the number of counted fish are arranged in chronological order, and a counting result including an image with a counted mark indicating fish that have been counted. A result providing unit that provides a display, and a correction unit that receives a correction operation and corrects the number of fish.

Miscounts are likely to occur when many fish pass at once and when many fish are counted per unit time. This configuration provides a fish count change display in which the display corresponding to the number of fish counted per unit time is arranged in chronological order. As a result, it is possible to grasp the point in time when a large number of fish are likely to pass at once, and erroneous counting is likely to occur, without visually observing the images for the entire time period. Then, the counting result display including the image with the counted mark indicating the counted fish can be visually confirmed and corrected through the correction operation. Therefore, erroneous counting can be easily corrected.

1 is a block diagram showing a fish counting system of this embodiment; FIG. It is a figure which shows the image|photographed image. FIG. 10 is a diagram showing a screen including counting result display and fish count change display; 4 is a flowchart showing an example of processing executed by the system; FIG. 10 is a diagram showing a screen including an image to which manual correction marks have been added; FIG. 10 is a diagram schematically showing another configuration of the fish counting system; FIG. 10 is a diagram showing another example of a screen including counting result display and fish count change display;

Embodiments of the present disclosure will be described below with reference to the drawings.

As shown in FIG. 1, the fish counting system 3 counts the number of fish based on a plurality of images, provides the display of the counting results to the provider of the display 4 or the like, and performs a correction operation via the operation unit 5. It is configured so that the number of accepted fish can be corrected. A plurality of images are obtained by capturing, with time, an imaging region in which the fluid containing the fish 1 flows with the camera 2 . FIG. 1 is an example, in which a first cage and a second cage are directly connected to move fish such as tuna from the first cage to the second cage. By arranging a camera near the entrance of the second cage, the number of fish that have moved to the second cage is counted, and the total number of fish is grasped. Note that the installation direction of the camera is not limited to the example shown in FIG. For example, it may be installed at the exit of the first fish tank.

In the present embodiment, the fish counting system 3 is implemented in a computer to which the display 4 and the operation unit 5 (devices such as a touch panel and mouse) are connected, but is not limited to this. For example, the fish counting system 3 may be realized by a server or cloud, and the display 4 and operation unit 5 of a computer that can communicate with the server or cloud may be used.

FIG. 2 is a diagram showing a photographed image. The direction of the camera 2 is installed in a predetermined direction. As for the predetermined orientation, as shown in FIG. 2, a part of the edge of the image is set as the first area Ar1, the other edge is set as the second area Ar2, and the first area Ar1 is close to the first cage. By arranging the camera 2 in such a predetermined direction, fish moving from the first cage to the second cage always pass through the first area Ar1 before reaching the second area Ar2. In FIG. 2, one of the four sides is set as the first area Ar1, and the remaining three sides are set as the second area Ar2, but the present invention is not limited to this.

In addition, since fish often pass through the center of the image, the shooting area has a third area Ar3 between the first area Ar1 and the second area Ar2. The third area Ar3 is arranged in the central part of the image. With such an arrangement, the fish 1 that has left the first area Ar1 does not immediately reach the second area Ar2, and the fish appears in the third area Ar3 in a plurality of images, thereby improving the recognition accuracy. becomes possible. Note that the third area Ar3 can be omitted. As shown in FIG. 1, the area setting data D1 regarding the first area Ar1, the second area Ar2 and the third area Ar3 are stored in the storage unit 39 in the system 3. FIG. The area setting data D1 may be set with an initial value, or may be changed separately by a user's operation. In this embodiment, the area setting data D1 is one file, and one area setting data D1 (one file) is stored corresponding to one moving image D2.

The fish counting system 3, as shown in FIG. have. Each of these units 30 to 38 is realized through cooperation of hardware and software by executing a predetermined program on one or more processors on a computer. The storage unit 39 is a memory, storage, or the like.

The image acquisition unit 30 shown in FIG. 1 acquires a plurality of images of an imaging region in which a fluid containing a fish 1 flows over time. The camera 2 shoots a moving image at, for example, 30 fps (Flames Per Second), and the moving image D2 is stored in the storage unit 39 of the system 3 . The image acquisition unit 30 acquires a plurality of images from the moving image D2. Multiple images are arranged in chronological order. In this specification, one image may be referred to as one frame.

The counting unit 34 shown in FIG. 1 counts the number of fish based on a plurality of images. The counting unit 34 is configured to count (add) the fish 1 when the fish 1 in the first area Ar1 moves to the second area Ar2. Specifically, the counting unit 34 extracts a fish by image processing for each of a plurality of images arranged in time series, and performs identification of an individual for the extracted fish based on the position of the fish in the past image. An individual identification label is given, and when a fish with the same label as the label given to fish 1 in the first area Ar1 reaches the second area Ar2, the fish are counted in the image (frame) at that time. This processing is an example, and various algorithms can be adopted as long as the number of fish can be counted.

As shown in FIG. 1, the counting unit 34 stores automatic counting data D3 indicating the counting result in the storage unit 39. As shown in FIG. In this embodiment, the automatic counting data D3 is one file, and one automatic counting data D3 (one file) is stored corresponding to one moving image D2. In this embodiment, the automatic counting data D3 has the number of fish counted (added) for each frame (each image). In the example shown in FIG. 1, the automatic counting data D3 indicates that the number of fish counted in the j-th frame is +3, the number of fish counted in the j+1-th frame is +2, and the number of fish counted in the j+2-th frame is +1, indicating that the number of fish counted in the j+3th frame is +4. Although j is a natural number, it is used for convenience of explanation.

The counting unit 34 also stores mark data D4 indicating the positions of the counted marks 61 (see FIG. 3) indicating fish that have been counted for each frame in the storage unit 39 (see FIG. 1). The mark data D4 is used to display the counted marks 61 (see FIG. 3) that indicate counted fish in the image. In this embodiment, the mark is a ROI (Region of Interest) display, and a rectangular frame is displayed superimposed on the image of the fish. The fish to be counted that have passed through the first area Ar1 but have not reached the second area Ar2 are represented by the counting target mark 68. FIG. The counted marks 61 and the counting target marks 68 are of different colors.

FIG. 3 shows an example of a screen including a counting result display 60. As shown in FIG. The result providing unit 36 shown in FIG. 1 provides a count result display 60 including an image 62 with a counted mark 61 indicating fish that have been counted, as shown in FIG. The counted mark 61 is a rectangular frame superimposed on the image of the fish indicated by the dashed line in FIG. 3, but is not limited to this. Various changes are possible as long as it is a ROI (Region of Interest) display. The counting result display 60 shown in FIG. 3 includes a count value 63 for the entire moving image (12,345 tails in the figure), a correction value 64 for the entire moving image (+18 tails in the figure), and an addition for inputting a correction operation. It includes an instruction display 65, a subtraction instruction display 66, and a fish count correction value 67 for the current frame. A count value 63 for the entire moving image is obtained by aggregating the automatic count data D3 shown in FIG. In addition, on this screen, you can play a video (multiple images), pause, advance the screen in units of frames, return the screen in units of frames, display frames with a specified time, and change the playback position with a specified time. is possible.

The correction unit 37 shown in FIG. 1 receives a correction operation and corrects the number of fish. In this embodiment, the correcting section 37 corrects the number of fish according to the operation on the addition instruction display 65 or the subtraction instruction display 66 . Specifically, when the addition instruction display 65 is operated, the fish number correction value 67 for the currently displayed image 62 (frame) is increased by +1. The correction value 67 corresponds to the number of times the addition instruction display 65 is operated. For example, if the addition instruction display 65 is pressed p times, the correction value 67 becomes +p. p is a natural number of 1 or more. Similarly, when the subtraction instruction display 66 is operated, the fish number correction value 67 for the currently displayed image (frame) is decremented by -1. The correction value 67 corresponds to the number of times the subtraction instruction display 66 is operated. For example, if the subtraction instruction display 66 is pressed p times, the correction value 67 becomes -p. Although p is a natural number, it is used for convenience of explanation.

The correction result (correction value 67) by the correction unit 37 is stored in the storage unit 39 as the manual correction data D5 shown in FIG. In this embodiment, the manual correction data D5 is one file, and one manual correction data D5 (one file) is saved corresponding to one moving image D2. In this embodiment, the manual correction data D5 has a correction value for each frame (each image). In the example shown in FIG. 1, the manual correction data D5 has a manually operated correction value of +10 in the (j+10)th frame, a correction value of +10 in the (j+100)th frame, and (j+320). It indicates that the correction value in the th frame is -2. The correction value 64 for the entire moving image displayed on the screen shown in FIG. 3 is obtained by totaling the manual correction data D5.

As shown in FIG. 1, the automatic count data D3 represents the number of fish counted by the counting unit 34. As shown in FIG. Manual correction data D5 represents the number of fish corrected by the correction unit 37 . The manual correction data D5 and the automatic counting data D3 are different data. The correction unit 37 changes the manual correction data D5 without changing the automatic counting data D3.

The fish count change display providing unit 38 shown in FIG. 1 provides the fish count change display 70 shown in FIG. The change display 70 has a display 71 corresponding to the number of fish counted per unit time, and the displays 71 are arranged in chronological order. In this embodiment, the fish number change display 70 has a graph display 71 as shown in FIG. In the graph display 71, a vertically extending first axis Ax1 corresponds to the number of fish counted per unit time, and a second axis Ax2 perpendicular to the first axis Ax1 corresponds to the passage of time. In the example of FIG. 3, the left represents the past and the right represents the future along the second axis Ax2. The number of fish counted per unit time increases from bottom to top along the first axis Ax1. When the shooting time of the moving image D2 is too long to be displayed on one screen, a slider bar 72 is displayed, and by sliding the slider bar 72, an arbitrary time point can be displayed. Further, by operating a time point to be visually recognized in the change display 70, the frame at the specified time point is displayed. Although the first axis Ax1 is vertical in this embodiment, it may be horizontal. In this embodiment, since the count value is managed for each frame, the unit time is the time corresponding to one frame. At 30 fps, it is 1/30 second. Of course, this is only an example, and the time corresponding to multiple frames may be treated as the unit time.

In the example shown in FIG. 3, the number of fish counted per unit time is indicated by a line graph, but this is only an example and various changes are possible. For example, various graphs such as bar graphs may be employed. Also, numbers as count values may be arranged in chronological order.

In this way, since the display 71 corresponding to the number of counted fish is arranged in chronological order, even if the images of all the times are not visually recognized, by looking at the display 71, erroneous counting is likely to occur. (the time when the count value is large) can be grasped. Further, in the case of the graph display 71, since the peak of the graph and its surroundings mean the point in time when the count value is large, it is possible to easily grasp the point in time when the count value is large.

The fish count change display 70 shown in FIG. 3 has a current display 73 representing the time point of the currently displayed image 62 (frame). The fish count change display 70 also has a corrected time point display 74 that indicates the time point corrected by the addition instruction display 65 or the subtraction instruction display 66 . The correction time point display 74 indicates not only the time point but also the content of the correction value (addition or subtraction).

<Operation flow>
Processing executed by the system 3 will be described with reference to FIG. A moving image D2 is stored in the storage unit 39 . When a predetermined analysis start condition is established, such as opening a predetermined screen, performing an analysis instruction operation, or storing a new moving image D2 in the storage unit 39, in step ST1 shown in FIG. , the image acquisition unit 30 acquires a plurality of images of the fluid containing the fish over time based on the moving image D2. At step ST2, the counting section 34 counts the number of fish based on the plurality of images. The counting result is stored in the storage section 39 as automatic counting data D3 and mark data D4 shown in FIG. When the fish number count/analysis screen shown in FIG. 3 is opened, in the next step ST3, the fish number change display providing unit 38 counts per unit time based on the automatic count data D3 and the mark data D4 shown in FIG. A fish count change display 70 is provided in which displays 71 corresponding to the number of fish caught are arranged in chronological order (see FIG. 3). Also, in step ST4, the result providing unit 36 provides a count result display 60 including an image 62 with a counted mark 61 indicating fish that have been counted, as shown in FIG. Step ST4 may be repeatedly executed according to the reproduction of the moving image D2 or the designation of the time point in the fish number change display 70, or the like. When the addition instruction display 65 or the subtraction instruction display 66 shown in FIG. 3 is operated, in step ST5, the correction section 37 accepts the correction operation and corrects the number of fish. The correction unit 37 stores the correction result in the storage unit 39 as manual correction data D5.

[Modification 1]
In the embodiment shown in FIG. 3, the number of fish is corrected for the displayed image 62 (frame) simply by operating the addition instruction display 65 or the subtraction instruction display 66, but the present invention is not limited to this. As a time-consuming but accurate correction method, the position of the fish that has been erroneously counted or should be counted in the image 62 may be designated and subjected to addition correction or subtraction correction. As a method of specifying the position of the fish, there are a method of specifying the position by enclosing it with a rectangular frame, a method of specifying coordinates within the image 62, and the like. It is preferable that the result providing unit 36 provides an image to which the counted mark 61 and the manual correction mark indicating the corrected fish designated by the correction operation are added. In this way, the grounds for the correction are displayed together with the image 62 using the manual correction mark, so appropriate fish quantity management is possible. FIG. 5 shows an example of a screen including an image 62 to which manual correction marks M have been added.

[Modification 2]
FIG. 6 schematically shows another configuration of the fish counting system 3 of this embodiment. The fish counting system 3 may further include a disturbance detector 31 in addition to the configuration of FIG. As with the image acquisition unit 30 and the like, the disturbance detection unit 31 is also implemented by cooperation of hardware and software by executing a predetermined program on one or more processors on a computer.

The disturbance detection unit 31 detects disturbance based on the image acquired by the image acquisition unit 30 . Then, the disturbance detection unit 31 provides a marking display 75 when the disturbance is detected in the fish number change display 70 arranged in chronological order (see FIG. 7). At this time, in addition to the marking display 75, the disturbance detection unit 31 may provide a disturbance content display 76 (see FIG. 7) indicating the detection content of the disturbance.

Here, the disturbance detection unit 31 detects disturbance by the following method for each type of disturbance (evaluation item).

When "reflection of direct sunlight" is detected as a disturbance (when the evaluation item is "reflection of direct sunlight"), the disturbance detection unit 31 divides the image acquired by the image acquisition unit 30 into blocks. Specifically, the image obtained by the image obtaining unit 30 is divided into 23×23 blocks while shifting the small area of 1/12 the size of the screen vertically and horizontally by 1/24 of the vertical and horizontal length of the screen. There is a block whose average brightness is above a certain value, whose average hue is below a certain value, and whose average value of each of R (red), G (green), and B (blue) is above a threshold. In this case, the disturbance detection unit 31 determines that there is reflection of direct sunlight (detects presence of disturbance).

When detecting “reflection of an obstacle such as a net or a structure” as disturbance (when the evaluation item is “reflection of an obstacle”), the disturbance detection unit 31 detects the image acquired by the image acquisition unit 30. , is divided into 4×4 blocks at equal intervals. Then, first, a block whose average brightness is equal to or less than a certain value is set as an obstacle candidate area. Next, a block adjacent to the obstacle candidate area and outside the obstacle candidate area is set as a "boundary with sea surface" candidate. Then, when more than half of the blocks adjacent to the obstacle candidate area exist in each block of the "boundary with the sea surface" candidate block whose brightness variance is equal to or greater than a certain value, the disturbance detection unit 31 detects the difference between the obstacle and the sea surface. exists and an obstacle is reflected (detects the presence of disturbance).

When detecting “reflection of floating objects” as a disturbance (when the evaluation item is “reflection of floating objects”), the disturbance detection unit 31 displays the image acquired by the image acquisition unit 30 in gray over the entire screen. After scaling, each pixel is binarized to perform contour extraction, and the area of each region surrounded by the contour is calculated. When the number of regions whose area falls within a certain range is equal to or greater than a certain value, the disturbance detection unit 31 determines that a floating object or the like is reflected (detects the presence of disturbance).

When detecting the "overall brightness and darkness of the image" as the disturbance (when the evaluation item is "overall brightness and darkness"), the brightness of the entire screen of the image acquired by the image acquiring unit 30 is equal to or greater than a certain value, the disturbance detection unit 31 determines that the entire image is too bright (detects presence of disturbance). Conversely, when the brightness of the entire screen of the image acquired by the image acquisition unit 30 is less than a certain value, the disturbance detection unit 31 determines that the entire image is too dark (detects the presence of disturbance).

FIG. 7 shows another example of a screen including a counting result display 60 and a fish number change display 70. FIG. The figure shows a state in which a marking display 75 is added to the fish number change display 70 at the time when the disturbance detection unit 31 detects a disturbance. If there is a disturbance such as reflection of direct sunlight, there is a possibility that the counting unit 34 cannot correctly count the number of fish based on the image. In the fish count change display 70 displayed in chronological order, marking display 75 is attached at the time when the disturbance detection unit 31 detects a disturbance, so that the erroneous counting caused by the disturbance is corrected according to the correction operation based on the marking display 75. becomes possible.

Further, in FIG. 7, in addition to the marking display 75, a disturbance content display 65 indicating the content of the disturbance detected by the above detection method is added to the fish number change display 70. FIG. Due to the presence of the disturbance content display 76, it becomes possible to perform correction according to the content of the disturbance when correcting erroneous counting caused by the disturbance (the degree of correction can be changed according to the content of the disturbance). , it is possible to accurately correct erroneous counts according to disturbances.

As described above, the fish counting system 3 of the present embodiment includes an image acquisition unit 30 that acquires a plurality of images of a fluid containing fish over time, and a counter that counts the number of fish based on the plurality of images. a fish count change display providing unit 38 that provides a fish count change display 70 in which a display 71 corresponding to the number of fish counted per unit time is arranged in chronological order; and a count indicating the counted fish. It includes a result providing unit 36 that provides a counting result display 60 including an image 62 with a completed mark 61 attached, and a correction unit 37 that receives a correction operation and corrects the number of fish.

The fish counting method of the present embodiment includes acquiring a plurality of images of a fluid containing fish over time, counting the number of fish based on the plurality of images, and counting the number of fish per unit time. providing a fish count change display in which displays corresponding to the number of fish are arranged in chronological order; providing a count result display including an image with a counted mark indicating the counted fish; receiving a correction operation to correct the number of fish.

Miscounts are likely to occur when many fish pass at once and when many fish are counted per unit time. This configuration provides a fish count change display 70 in which displays 71 corresponding to the number of fish counted per unit time are arranged in chronological order. As a result, it is possible to grasp the point in time when a large number of fish are likely to pass at once, and erroneous counting is likely to occur, without visually observing the images for the entire time period. Then, the counting result display 60 including the image with the counted mark 61 indicating the fish that has been counted can be visually confirmed and corrected through the correction operation. Therefore, erroneous counting can be easily corrected.

As in this embodiment, the fish number change display 70 corresponds to the number of fish counted per unit time on the first horizontal or vertical axis Ax1, and the second axis Ax2 perpendicular to the first axis Ax1 indicates the elapsed time. It is preferred to have a graphical representation 71 corresponding to .

By looking at the graph display 71, it becomes possible to easily grasp the point in time when counting errors tend to occur by looking at the change in the graph display 71, which facilitates correction. Time points at which erroneous counting is likely to occur include, for example, peaks and sudden rises.

As in the present embodiment, the result providing unit 36 displays an image 62 with a counted mark 61 indicating fish that have been counted, an addition instruction display 65 and a subtraction instruction display 66 for instructing correction of the number of fish. It is preferable that the correction unit 37 corrects the number of fish according to the operation on the addition instruction display 65 or the subtraction instruction display 66 .

According to this configuration, when visually observing the image 62 with the counted mark 61 attached and recognizing an omission of counting without the counted mark 61 attached or an erroneous count with the counted mark 61 erroneously attached, Since the addition instruction display 65 or the subtraction instruction display 66 can be operated, the efficiency of correction work can be improved.

As in Modification 1 of the present embodiment, the correction unit 37 accepts a correction operation that involves specifying the positions of corrected fish in the image 62, corrects the number of fish, and the result providing unit 36 specifies by the correction operation. provides an image 62 with a manual correction mark M indicating the corrected corrected fish.

In this way, the grounds for the correction are displayed together with the image 62 by the manual correction mark M, so appropriate fish count management becomes possible.

As in the present embodiment, the fish count change display 70 preferably has a corrected time point display 74 indicating the time point when the fish count was corrected.

According to this configuration, the display 71 corresponding to the number of fish counted per unit time and the correction time display 74 are displayed. It is possible to improve convenience. Even so, when a plurality of people corrects the erroneous count, the person who performed the correction first can easily grasp the corrected portion, and efficient checking and correction work by a plurality of people is possible.

As in the present embodiment, the fish counting system 3 has a storage unit 39, and the storage unit 39 contains automatic counting data D3 representing the number of fish counted by the counting unit 34, and data different from the automatic counting data D3. Manual correction data D5 representing the number of fish corrected by the correction unit 37 is stored, and the correction unit 37 preferably changes the manual correction data D5 without changing the automatic counting data D3.

With this configuration, the correction does not damage the automatic count data D3, so even if the correction is made by mistake, it can be easily restored.

As in Modification 2 of the present embodiment, the fish counting system 3 further includes a disturbance detection unit 31 that detects disturbance based on the image acquired by the image acquisition unit 30, and the fish number change display 70 is a disturbance detection unit. Preferably, 31 has a marking indicia 75 to indicate when the disturbance has been detected. In this case, according to the correction operation based on the marking display 75, it is possible to correct the erroneous counting caused by the disturbance.

Moreover, the fish count change display 70 preferably has a disturbance content display 76 indicating the detection content of the disturbance. In this case, according to the correction operation based on the disturbance content display 76, correction according to the content of the disturbance can be performed. This makes it possible to accurately correct erroneous counts caused by disturbances.

A program according to this embodiment is a program that causes one or more processors to execute the above method. Further, the computer-readable temporary recording medium according to the present embodiment stores the above program.

Each unit 30 to 38 shown in FIG. 1 is realized by executing a predetermined program by one or a processor, but each unit may be configured by a dedicated memory or a dedicated circuit.

In the system 3 of the above embodiment, the units 30 to 38 are realized by computer processors, but the units 30 to 38 may be distributed and implemented in a plurality of computers or in the cloud. That is, the method may be performed by one or more processors.

Although the embodiments of the present disclosure have been described above based on the drawings, it should be considered that the specific configurations are not limited to these embodiments. The scope of the present disclosure is indicated not only by the description of the above embodiments but also by the scope of claims, and includes all modifications within the meaning and scope equivalent to the scope of claims.

INDUSTRIAL APPLICABILITY The present invention is applicable to fish counting systems.

1 fish 3 fish counting system 30 image acquisition unit 34 counting unit 36 result providing unit 37 correcting unit 38 fish number change display providing unit 39 storage unit 61 counted mark 65 addition instruction display 66 subtraction instruction display 70 fish number change display 71 graph display 74 Correction time display Ax1 1st axis Ax2 2nd axis D3 Automatic counting data D5 Manual correction data

Claims (10)

an image acquisition unit that acquires a plurality of images of a fluid containing fish over time;
a counting unit that counts the number of fish based on the plurality of images;
a fish count change display providing unit that provides a fish count change display in which displays corresponding to the number of fish counted per unit time are arranged in chronological order;
a result providing unit providing a counting result display including an image with a counted mark indicating fish that have been counted;
a correction unit that receives a correction operation and corrects the number of fish. The fish number change display has a graph display in which a first axis extending horizontally or vertically corresponds to the number of fish counted per unit time, and a second axis orthogonal to the first axis corresponds to the passage of time. Item 2. The fish counting system according to Item 1. The result providing unit provides an image with a counted mark indicating fish that have been measured, and an addition instruction display and a subtraction instruction display for instructing correction of the number of fish,
3. The fish counting system according to claim 1, wherein said correction unit corrects the number of fish according to an operation to said addition instruction display or said subtraction instruction display. The correction unit receives a correction operation that specifies the position of the fish to be corrected in the image, corrects the number of the fish,
4. The fish counting system according to any one of claims 1 to 3, wherein said result providing unit provides an image to which a manual correction mark indicating said fish to be corrected designated by said correction operation is added. 5. The fish counting system according to any one of claims 1 to 4, wherein said fish count change display has a correction time display indicating a time when said fish count is corrected. Equipped with a storage unit,
The storage unit stores automatic counting data representing the number of fish counted by the counting unit, and manual correction data representing the number of fish corrected by the correction unit, which is different from the automatic counting data, and
The fish counting system according to any one of claims 1 to 5, wherein said correction unit changes said manual correction data without changing said automatic counting data. Further comprising a disturbance detection unit that detects a disturbance based on the image acquired by the image acquisition unit,
7. The fish counting system according to any one of claims 1 to 6, wherein said fish count change display has a marking display indicating a time point at which said disturbance detection section detects said disturbance. 8. The fish counting system according to claim 7, wherein said fish number change display has a disturbance content display indicating detection content of said disturbance. obtaining a plurality of images over time of a fluid containing fish;
counting the number of fish based on the plurality of images;
providing a fish count change display in which displays corresponding to the number of fish counted per unit time are arranged in chronological order;
providing a count results display including an image with a counted mark showing fish that have been counted;
receiving a correction operation to correct the number of fish. A program that causes one or more processors to execute the fish counting method according to claim 9.

Images